Apple discontinued the $599 Mac mini earlier this year. Not because it was a bad product. Because demand outpaced supply, and a lot of that demand came from people building AI agent setups on local hardware.
That's worth paying attention to.
For years, the assumption was that AI meant cloud. You send your data to a server somewhere, the server does the thinking, you get an answer back. And for most businesses, that still works fine. But something shifted.
Local AI on a Mac mini, a dedicated workstation, or even a high-end laptop is now a practical option. Not a hobbyist project. A real alternative you can evaluate against what you're already paying for.
So what's the actual difference?
Cloud AI runs on someone else's infrastructure. You pay per use, you get near-instant access to the latest models, and you don't have to manage anything yourself. For most workflows, especially ones that are bursty or unpredictable, cloud is the right call. The cost is low when you're not using it, and you're not buying hardware.
Local AI runs on hardware you own. The model lives on your machine. Once it's set up, it runs offline, it doesn't send your data anywhere, and there's no per-query cost. If you have workflows that run constantly, handle sensitive information, or need fast response times without network latency, local starts making real economic and practical sense.
Here's how a small business might think about it.
A law firm handling sensitive client intake probably doesn't want that data hitting a third-party API every time someone fills out a form. Running a local model for initial screening is cleaner, keeps data in-house, and costs nothing per query.
A marketing agency generating dozens of client reports every month might find cloud costs creeping up as usage scales. A local setup with a solid model handles that workload at a flat hardware cost.
A solo consultant who travels and works offline benefits from a model that doesn't need a connection to function.
None of this means cloud AI is going away. It's not. For most businesses, cloud is still the starting point because the setup is minimal and the models are powerful. But the gap between cloud-only and local-capable is closing fast, and hardware like the Mac mini made that accessible at a price most small businesses can justify.
The question isn't AI or no AI. It's figuring out which setup fits your business.
If you're running workflows that are consistent, sensitive, or high-volume, local is worth a conversation. If you're experimenting or handling unpredictable workloads, cloud is the right starting point.
Most businesses end up using both. Cloud for the flexible, fast-moving stuff. Local for the steady, sensitive, or cost-sensitive workflows.
The Mac mini selling out wasn't a trend. It was a signal. Local AI is no longer a fringe setup. It's a real option, and more businesses are figuring that out.